Application of Collaborative Filtering Recommendation Algorithm in Internet Online Courses

Abstract

Aiming at the problem that the overload of online education platform course resources leads to the difficulty of user selection, this paper mainly studies the improvement and application of collaborative filtering algorithm based on online course recommendation system, which organically combines personalized recommendation technology and online course system to meet the needs of users and online education platform. In the process of recommendation, firstly, user preferences are collected to establish a data model, and user login information and learning behavior information are used as implicit characteristics of user preferences. The loss rate of users in the computing platform is defined, the popularity of each course is calculated, and the relationship between users and courses is constructed, and the correlation and comparative analysis are carried out, Then, the traditional collaborative filtering algorithm is improved by introducing the implicit features after analysis, and the cosine similarity method is used to calculate the course similarity. Finally, the topN recommendation list is generated to get the recommendation results. Based on the desensitization data of an education platform, the experimental results show that the improved recommendation model can improve the precision of recommendation by introducing implicit features.